Browsing by Author "Clark, William Walker"

Now showing 1 - 2 of 2
  • Thumbnail Image
    Application of adaptive trusses to vibration isolation in flexible structures
    Clark, William Walker (Virginia Tech, 1991-12-05)
    This dissertation presents techniques for using adaptive trusses for active vibration isolation in flexible structures. Passive methods have been used almost exclusively in the past for vibration isolation, although in the more recent literature active techniques have been proposed in an attempt to achieve greater isolation performance. Most of the active techniques, however, require either detailed knowledge of the system or of the disturbance to be isolated. This work focuses on techniques in which knowledge of the disturbance is minimal, and in some cases, knowledge of the system is not necessary. Two new active vibration isolation methods are presented which are based on feedback of transmitted forces in the system. The methods include force feedback through a high gain, and state feedback using the LQR method with disturbance modelling. A third method which has been demonstrated in the literature, force feedback through a classical compensator, is also presented for comparison. For the purpose of discussion, each of the methods is applied to a system which includes a single active mount. The methods are then applied analytically to an adaptive truss, which essentially contains multiple mounts, to demonstrate multi-degree-of-freedom active vibration isolation. It is shown that force feedback provides two-way isolation, and its effects are independent of the type of active mount used (whether it is a force- or displacement-commanded mount). The most promising technique proves to be the simplest, the high-gain feedback method. This technique is a stable, model-free method of vibration isolation which places no restrictions on the type of system disturbance, other than that it must be within the actuator’s bandwidth. The high-gain approach is applied experimentally and shown to agree with the simulated results.
  • Thumbnail Image
    A planar comparison of actuators for vibration control of flexible structures
    Clark, William Walker (Virginia Tech, 1988-05-12)
    Interest in large flexible space structures has grown considerably over the last decade. These distributed parameter systems exhibit vibration characteristics such as low, closely spaced natural frequencies and light damping, which, when coupled with the stringent pointing accuracy and vibration control requirements imposed on these systems, bring about interesting control problems. Addressing these problems has called for the use of active vibration control. Up to now, two of the most popular means for active vibration control of large space structures have been proof mass and reaction wheel actuators. These actuators are inertial-type actuators in that they operate by applying forces or moments to masses whose reaction forces, imposed on the structure, act to dampen the vibrations of the structure. A new class of actuators, variable geometry trusses (VGT's), has been recently introduced. These actuators are actually built into the structure, and they operate by varying their link lengths to apply forces to the structure or to change the shape of the structure itself. This study compared the effectiveness of four actuators in controlling the planar vibrations of a cantilevered truss-beam. The actuators chosen for the study were a proof mass actuator, a reaction wheel actuator, and two VGT's, the planar truss actuator, and the planar truss proof mass actuator (a combination VGT/inertial type actuator). Numerical simulations of each beam/actuator system were performed in response to initial condition inputs. A full-state, LQR optimal feedback control law was used with each system. These simulations provided information such as time response of the closed-loop system, damping provided to the beam, and power required by each actuator. This information can be used to determine the "best" actuator for a given purpose. The results of these simulations show that the VGT's are preferable in terms of damping added to the beam. The proof mass actuator is more efficient as far as power required to do the control, however, the efficiencies for all actuators are very similar.